OpenAI's GPT-6 Cyber found 81 Firefox flaws and exploits for 948 Linux kernel bugs
OpenAI says the model is its first rated "Critical" for cyber capability under its Preparedness Framework, and it isn't publicly available yet.
OpenAI announced GPT-6 Cyber on Thursday, and the numbers it published are the sort that usually come from a research team, not a model release. During security testing the model found 81 previously unknown high-severity vulnerabilities in Firefox and helped develop patches, many of which Mozilla has already released. In separate Linux kernel work it flagged 2,308 potential flaws, reproduced 1,724 of them, and built local privilege escalation exploits for 948 distinct vulnerabilities that could let an unprivileged user get root.
On OpenAI's ExploitGym benchmark it exploited 48.5% of vulnerabilities, against 32.1% for GPT-5.6 Cyber. OpenAI has rated it "Critical" for cybersecurity capability under its Preparedness Framework, up from "High" for the previous model.
You can't use it. Access is restricted to approved customers in Daybreak Red, which supports exploit development and authorized testing; OpenAI says it plans to widen that soon. Daybreak Blue, for verified defenders, has fewer restrictions.
OpenAI's own framing is that defenders have a limited window to use frontier models to find and patch bugs before similar capability reaches attackers. We'd read the Firefox line as the useful one: patches exist, so check what your browser is running.
The advisory doesn't say how many of the 81 fixes have shipped, or how the Linux kernel findings were reported upstream.
Sources
- OpenAI's new GPT-6 Cyber AI finds 81 Firefox flaws, 948 Linux bugsPrimarywww.bleepingcomputer.com
If something here is wrong, tell us. Every correction is listed on the corrections page.
Don’t Just Read About Cyberattacks. Learn How They Work.
Go beyond the headlines. Explore hands-on cybersecurity labs, understand real-world vulnerabilities, and build practical security skills with HackZoneLabs.
Start Learning
